Self Organized Physical Cell ID Assignment in Multi-Operator Heterogeneous Networks

We discuss Physical Cell ID (PCI) assignment for densely deployed heterogeneous networks (HetNets), in a multioperator spectrum sharing scenario. The aim is to achieve PCI assignment that is conflict-free and confusion-free, jointly for multiple operators, in a self-organized way. A graph coloring formulation is used to model the problem where the PCIs represent the colors used to color the interference graph. The graph is created using the interference couplings determined by handover measurements of the user equipments served by the HetNet. The PCI assignment algorithms considered are based on local search methods used in conjunction with the focused search principle, which allows PCI reconfigurations only to the cells with unsatisfied constraints related to conflicts and confusions. The algorithms considered are fully distributed and do not involve message-passing between the operators. Each cell decides its PCI by observing the PCIs of its neighbors which include its own network cells as well as the cells that belong to the other operators but share the same spectrum. Simulations results show that the proposed algorithms have good convergence properties and are effective for self-organized PCI assignment in dense multioperator networks with shared spectrum.